JPWO2015194235A1 - Sampling data processing apparatus, sampling data processing method, and computer program - Google Patents

Abstract

Abstract: A sampling data processing apparatus, method, and computer program capable of accurately estimating the distribution of a product lot by estimating the distribution of product lots outside the product standard from sampling data of product lots. provide. An input of an upper limit value and a lower limit value of a predetermined characteristic value based on the product standard of the target product is received, and an average value of the standard deviation of the characteristic value is calculated based on the control chart of the product lot. An average value of the characteristic values is calculated from the sampled measurement data, and an upper limit value and a lower limit value of the average value of the characteristic values in the confidence interval of 95% reliability are calculated. Estimate the measurement standard deviation that shows the variation of the measuring instrument itself with respect to the characteristic value, and the average value of the calculated characteristic value is the average value of the characteristic value of the confidence interval corresponding to the closer of the upper limit value and lower limit value The upper limit value or lower limit value is updated as the average value of the characteristic values. The standard deviation of the characteristic value of the product itself is estimated, the upper defective rate and the lower defective rate are calculated, and the non-defective product rate is calculated.

Description

  The present invention relates to a sampling data processing apparatus, a sampling data processing method, and a computer program that estimate the quantity of product lots that are out of product standards from sampling data of product lots.

  A product is measured for a characteristic value indicating a predetermined characteristic before shipping, and is selected as a non-defective product or a defective product depending on whether or not a predetermined standard is satisfied. Product selection is performed by comparing the measured characteristic value of the product with an inspection standard whose conditions are stricter than the product standard (characteristic value required for the product). If the measured product characteristic value variation is only the characteristic value variation of the product itself, even if the inspection standard is stipulated under the same conditions as the product standard, the product is correctly selected as good or defective. be able to.

  However, the measured variation in the characteristic value of the product includes not only the variation in the characteristic value of the product itself but also the variation in the measured value of the measurement system. For this reason, there is a possibility that a defective product is included in a product selected as a non-defective product, or a non-defective product is included in a product selected as a defective product.

  In addition, it is difficult to select products for all products, and a sampling inspection is usually performed in which a certain amount of sampling is performed. In the sampling inspection, the smaller the average number of products to be inspected, the lower the cost of product inspection.

  For example, in Non-Patent Document 1, it is assumed that the quality characteristics follow a “normal distribution”, and a method for guaranteeing the average detection quality limit by guaranteeing that the lot acceptance probability of the lot allowable defect rate is not more than a specified value. A method is disclosed in which the average number of inspections can be significantly reduced as compared with the selective sampling inspection specified in the product standard.

"Design of weighing and sampling sampling inspection-sorting sampling inspection when quality characteristics follow a normal distribution", Japan Business Engineering Journal, Vol. 42, no. 6 (1992), p. 397-405

  The method disclosed in Non-Patent Document 1 is based on the premise that the product characteristics follow a “normal distribution”, and thus the accuracy of the OC curve is improved using the characteristics of the “normal distribution”. Therefore, the method is basically the same as the conventional OC curve estimation method, and the distribution of product lots is not specifically estimated, so that it is difficult to obtain sufficient estimation accuracy.

  The present invention has been made in view of the above circumstances, and specifically estimates the quantity of a product lot existing outside the product standard from the product lot sampling data, and the distribution of the product lot using the control chart data. Accordingly, an object of the present invention is to provide a sampling data processing device, a sampling data processing method, and a computer program that can be estimated with high accuracy.

  In order to achieve the above object, a sampling data processing apparatus according to the present invention is a sampling data processing apparatus for estimating the quantity of a product lot existing outside the product standard from sampling data of a product lot of the target product. A standard range receiving means for receiving an input of an upper limit value and a lower limit value of a predetermined characteristic value based on the standard; a standard deviation calculating means for calculating an average value of the standard deviation of the characteristic value based on a control chart of the product lot; Based on sample average calculation means for calculating an average value of the characteristic values from measurement data of a predetermined number of samples extracted from the product lot, based on the average value of the standard deviation of the calculated characteristic values and the average value of the characteristic values A confidence interval calculation means for calculating an upper limit value and a lower limit value of an average value of the characteristic values of a confidence interval of 95% confidence level, and a variation in the measuring instrument itself relating to the characteristic values. Measurement standard deviation calculating means for estimating the measurement standard deviation to be shown, and determining whether the calculated average value of the characteristic values is closer to the upper limit value or the lower limit value of the characteristic values that have been accepted, and corresponds to the closer one Based on the average value updating means for updating the upper limit value or lower limit value of the average value of the characteristic values of the confidence interval as the average value of the characteristic values, the average value of the standard deviation of the characteristic values, and the measurement standard deviation Product standard deviation estimating means for estimating the standard deviation of the characteristic value of the product itself, an upper defective rate that is a ratio of defective products exceeding an upper limit value of a predetermined characteristic value based on a product standard of the target product, and the target product A defective rate calculating means for calculating a lower defective rate that is a ratio of defective products that are lower than a lower limit value of a predetermined characteristic value based on a product standard, and a non-defective product rate based on the calculated upper defective rate and the lower defective rate Non-defective product rate to calculate Characterized in that it comprises left and means.

  Next, in order to achieve the above object, a sampling data processing apparatus according to the present invention is a sampling data processing apparatus that estimates the quantity of product lots that are out of product specifications from sampling data of product lots of target products. Standard range receiving means for receiving an input of an upper limit value and a lower limit value of a predetermined characteristic value based on the product standard of the product, and a standard deviation calculation for calculating an average value of the standard deviation of the characteristic value based on the control chart of the product lot Means, a measurement standard deviation calculating means for estimating a measurement standard deviation indicating a variation of the measuring device itself with respect to the characteristic value, an average value of the standard deviation of the characteristic value, and the measurement standard deviation based on the measurement standard deviation. An average for sequentially determining an average value of the characteristic values between a product standard deviation estimating means for estimating a standard deviation of the characteristic values and an upper limit value and a lower limit value of the characteristic values received. And an upper defective rate that is a ratio of defective products exceeding an upper limit value of a predetermined characteristic value based on a product standard of the target product, and a product standard of the target product for each average value of the specified characteristic values. Based on the calculated upper defect rate and the lower defect rate, the non-defective rate is successively calculated based on the calculated upper defect rate and the lower defect rate. A non-defective product rate calculating means for calculating the number of samplings of the corresponding sample by specifying a section of the average value of the characteristic values in which the calculated non-defective product rate is equal to or greater than a predetermined threshold. Features.

  Next, in order to achieve the above object, the sampling data processing method according to the present invention is executed by a sampling data processing apparatus that estimates the quantity of product lots existing outside the product standard from the sampling data of the product lot of the target product. In the sampling data processing method, the sampling data processing device receives an input of an upper limit value and a lower limit value of a predetermined characteristic value based on a product standard of the target product, and based on the control chart of the product lot A step of calculating an average value of the standard deviation of the characteristic value; a step of calculating an average value of the characteristic value from measurement data of a predetermined number of samples extracted from the product lot; and a standard deviation of the calculated characteristic value And calculating an upper limit value and a lower limit value of the average value of the characteristic value of the 95% confidence interval based on the average value of the characteristic value and the average value of the characteristic value, and The process of estimating the measurement standard deviation indicating the variation of the measuring device itself, and determining whether the calculated average value of the characteristic values is closer to the upper limit value or the lower limit value of the characteristic values that have been accepted. Updating the upper limit value or lower limit value of the average value of the characteristic values of the confidence interval corresponding to the average value of the characteristic values, the average value of the standard deviation of the characteristic values, and the measurement standard deviation , Based on the step of estimating the standard deviation of the characteristic value of the product itself, the upper defective rate that is the ratio of defective products exceeding the upper limit value of the predetermined characteristic value based on the product standard of the target product, and the product standard of the target product Including a step of calculating a lower defect rate that is a ratio of defective products that are below a lower limit value of a predetermined characteristic value, and a step of calculating a non-defective product rate based on the calculated upper defect rate and the lower defect rate. It is characterized by.

  Next, in order to achieve the above object, the sampling data processing method according to the present invention is executed by a sampling data processing apparatus that estimates the quantity of product lots existing outside the product standard from the sampling data of the product lot of the target product. In the sampling data processing method, the sampling data processing device receives an input of an upper limit value and a lower limit value of a predetermined characteristic value based on a product standard of the target product, and based on the control chart of the product lot A step of calculating an average value of the standard deviation of the characteristic value, a step of estimating a measurement standard deviation indicating variation of the measuring instrument itself with respect to the characteristic value, an average value of the standard deviation of the characteristic value, and the measurement standard deviation The average value of the characteristic values is sequentially determined between the step of estimating the standard deviation of the characteristic value of the product itself and the upper limit value and the lower limit value of the characteristic value for which input has been received. And an upper defective rate that is a ratio of defective products exceeding an upper limit value of a predetermined characteristic value based on a product standard of the target product, and a product standard of the target product for each of the specified average values of the characteristic values. A non-defective product rate is sequentially calculated based on the step of calculating a lower defect rate that is a ratio of defective products that are below a lower limit value of a predetermined characteristic value based on the upper limit defect rate and the lower defect rate. The number of samples to be extracted is calculated by specifying a section of the average value of the characteristic values in which the calculated non-defective product ratio is equal to or greater than a predetermined threshold.

  Next, in order to achieve the above object, the computer program according to the present invention may be executed by a sampling data processing apparatus that estimates the quantity of product lots that are outside the product standard from sampling data of product lots of the target product. In the possible computer program, the sampling data processing device is a standard range receiving means for receiving an input of an upper limit value and a lower limit value of a predetermined characteristic value based on a product standard of the target product, and the characteristic based on the control chart of the product lot A standard deviation calculating means for calculating an average value of standard deviations of values, a sample average calculating means for calculating an average value of the characteristic values from measurement data of a predetermined number of samples extracted from the product lot, Based on the average value of the standard deviation and the average value of the characteristic values, the upper limit value of the average value of the characteristic values in the confidence interval of 95% reliability A confidence interval calculation means for calculating a limit value, a measurement standard deviation calculation means for estimating a measurement standard deviation indicating variation of the measuring instrument itself with respect to the characteristic value, and an average value of the calculated characteristic value is an input of the characteristic value An average value updating unit that determines which one of the upper limit value and the lower limit value is closer to each other and updates the upper limit value or lower limit value of the average value of the characteristic values of the confidence interval corresponding to the closer one as the average value of the characteristic values Product standard deviation estimating means for estimating the standard deviation of the characteristic value of the product itself based on the average value of the standard deviation of the characteristic value and the measured standard deviation, and a predetermined characteristic value based on the product standard of the target product A defect rate calculation means for calculating an upper defect rate that is a ratio of defective products exceeding an upper limit value, and a lower defect rate that is a ratio of defective products that are lower than a lower limit value of a predetermined characteristic value based on a product standard of the target product; and Calculation Wherein the function as non-defective ratio calculating means for calculating the yield rate on the basis of the upper fraction defective and the lower failure rate.

  Next, in order to achieve the above object, the computer program according to the present invention may be executed by a sampling data processing apparatus that estimates the quantity of product lots that are outside the product standard from sampling data of product lots of the target product. In the possible computer program, the sampling data processing device is a standard range receiving means for receiving an input of an upper limit value and a lower limit value of a predetermined characteristic value based on a product standard of the target product, and the characteristic based on the control chart of the product lot A standard deviation calculating means for calculating an average value of standard deviations of values, a measurement standard deviation calculating means for estimating a measurement standard deviation indicating variation of the measuring instrument itself with respect to the characteristic values, an average value of the standard deviations of the characteristic values, and Product standard deviation estimating means for estimating the standard deviation of the characteristic value of the product itself based on the measurement standard deviation, the characteristic having received the input An average specifying means for sequentially specifying an average value of the characteristic values between the upper limit value and the lower limit value of the predetermined value, and for each of the specified average values of the characteristic values, a predetermined characteristic based on a product standard of the target product. Defective rate calculating means for calculating an upper defective rate that is a ratio of defective products that exceed the upper limit value and a lower defective rate that is a ratio of defective products that are lower than a lower limit value of a predetermined characteristic value based on the product standard of the target product A non-defective product rate calculating means for sequentially calculating a non-defective product rate based on the calculated upper defective rate and the lower defective rate, and an average value of the characteristic values in which the calculated non-defective product rate is equal to or greater than a predetermined threshold value. By specifying this section, it is made to function as a means for calculating the sampling number of the corresponding sample.

  The sampling data processing apparatus, sampling data processing method, and computer program according to the present invention can estimate the quantity of product lots outside the product standard from the sampling data of product lots with high accuracy by the above configuration. . On the contrary, by estimating the distribution of the non-defective product rate, it is possible to reversely calculate the number of samples to be extracted so as to become a predetermined non-defective product rate, for example, 99.99% or more.

It is a block diagram which shows the structural example of the sampling data processing apparatus which concerns on Embodiment 1 of this invention. It is a functional block diagram of the sampling data processing apparatus which concerns on Embodiment 1 of this invention. It is an illustration figure of the management chart of a product lot. It is a schematic diagram which shows the upper side failure rate and lower side failure rate of the sampling data processing apparatus which concerns on Embodiment 1 of this invention. It is a flowchart which shows the process sequence of CPU of the arithmetic processing part of the sampling data processing apparatus which concerns on Embodiment 1 of this invention. It is a functional block diagram of the sampling data processing apparatus which concerns on Embodiment 2 of this invention. It is a graph which shows the fluctuation | variation of the yield rate at the time of changing the variable i to 0-100. It is the graph which expanded and displayed the two area | regions of FIG. It is a flowchart which shows the process sequence of CPU of the arithmetic processing part of the sampling data processing apparatus which concerns on Embodiment 2 of this invention.

  Hereinafter, a sampling data processing apparatus for estimating the number of product lots existing outside the product standard from the sampling data of product lots in the embodiment of the present invention will be specifically described with reference to the drawings. The following embodiments do not limit the invention described in the claims, and all combinations of characteristic items described in the embodiments are essential to the solution. It goes without saying that it is not limited.

  In the following embodiments, a sampling data processing apparatus in which a computer program is introduced into a computer system will be described. However, as will be apparent to those skilled in the art, the present invention is a computer capable of executing a part thereof by a computer. Can be implemented as a program. Therefore, the present invention can take an embodiment of hardware as a sampling data processing device, an embodiment of software, or an embodiment of a combination of software and hardware. The computer program can be recorded on any computer-readable recording medium such as a hard disk, DVD, CD, optical storage device, magnetic storage device or the like.

(Embodiment 1)
FIG. 1 is a block diagram showing a configuration example of a sampling data processing apparatus according to Embodiment 1 of the present invention. The sampling data processing apparatus according to the first embodiment includes a measuring unit 1 that measures a characteristic value indicating a predetermined characteristic of a product, and an arithmetic processing unit 2 that calculates the measured characteristic value.

  The measuring unit 1 measures a characteristic value indicating a predetermined characteristic of the product. For example, when the product is a ceramic capacitor, the measurement unit 1 measures the capacitor capacity, which is a characteristic value of the product. As a hardware configuration of the measurement unit 1 that measures the capacitor capacity, there is an LCR meter.

  The arithmetic processing unit 2 is connected to at least a CPU (Central Processing Unit) 21, a memory 22, a storage device 23, an I / O interface 24, a video interface 25, a portable disk drive 26, a measurement interface 27, and the above-described hardware. The bus 28 is configured.

  The CPU 21 is connected to the above-described hardware units of the arithmetic processing unit 2 via the internal bus 28, and controls the operation of the above-described hardware units and stores the computer program 230 stored in the storage device 23. Various software functions are executed according to the above. The memory 22 is composed of a volatile memory such as SRAM or SDRAM, and a load module is expanded when the computer program 230 is executed, and stores temporary data generated when the computer program 230 is executed.

  The storage device 23 includes a built-in fixed storage device (hard disk), a ROM, and the like. The computer program 230 stored in the storage device 23 is downloaded by the portable disk drive 26 from a portable recording medium 90 such as a DVD or CD-ROM in which information such as programs and data is recorded. To the memory 22 and executed. Of course, it may be a computer program downloaded from an external computer connected to the network.

  The measurement interface 27 is connected to the internal bus 28, and by connecting to the measurement unit 1, it is possible to transmit / receive characteristic values and control signals measured between the measurement unit 1 and the arithmetic processing unit 2. It has become.

  The I / O interface 24 is connected to a data input medium such as a keyboard 241 and a mouse 242 and receives data input. The video interface 25 is connected to a display device 251 such as a CRT monitor or LCD, and displays a predetermined image.

  Hereinafter, the operation of the sampling data processing apparatus having the above-described configuration will be described. FIG. 2 is a functional block diagram of the sampling data processing apparatus according to Embodiment 1 of the present invention. The measuring unit 1 measures a characteristic value indicating a predetermined characteristic of the product 10.

  The standard range receiving unit 201 receives an input of an upper limit value and a lower limit value of predetermined characteristic values based on the product standard of the target product. For example, assuming a product lot consisting of 10,000 capacitors with a capacitance of 2.5 pF as a characteristic value, the upper limit product standard (upper limit: SUL) of the capacitor is 2.64 pF, and the lower limit product standard (lower limit) : SLL) is 2.44 pF. Therefore, the user inputs the upper limit value and the lower limit value of the predetermined characteristic value of the product via the keyboard 241 or the like. Of course, it may be stored in advance in the storage device 23 via a communication network (not shown).

  The sample average calculating means 202 calculates the average value of the characteristic values from the measurement data of a predetermined number of samples extracted from the product lot. For example, 10 products are extracted as samples from the product lot, and the capacitance of the sample extracted by the measurement unit (LCR meter) 1 is measured. An average value of the measured characteristic values is calculated, and for example, an average value Xbar (2.5388 pF) of the characteristic values is stored in the memory 22 or the like.

  The standard deviation calculation means 203 calculates the average value of the standard deviation of the characteristic values based on the product lot control chart. FIG. 3 is an exemplary diagram of a product lot management chart.

  As shown in FIG. 3, the standard deviation (pF) of each product lot is plotted for 30 product lots. Therefore, for example, an average value 0.020531 pF of standard deviations of 30 lots can be calculated as an average value σ sbar (pF) of standard deviations of overall characteristic values.

  Returning to FIG. 2, the confidence interval calculation means 205 is based on the stored average value Xbar (2.5388 pF) of the characteristic values and the average value σsbar (0.020531 pF) of the standard deviation of the calculated characteristic values. An upper limit value and a lower limit value of the average value of the characteristic values of the confidence interval of 95% reliability are calculated. Specifically, the upper limit value and the lower limit value of the average value of the characteristic values of the confidence interval of 95% reliability are calculated according to (Equation 1) and (Equation 2), respectively.

  However, the control chart coefficient C4 in (Expression 1) and (Expression 2) is calculated according to (Expression 3).

  For example, when the number of samples n is 10 in (Expression 3), n = 10 is substituted to obtain the control chart coefficient C4 = 0.97266. Substituting into (Equation 1) and (Equation 2) respectively, the upper limit value of the average value of the characteristic values of the confidence interval is 2.552150 pF, and the lower limit value is 2.525450 pF.

  Next, the measurement standard deviation calculation means 204 estimates the measurement standard deviation S1 of the characteristic value indicating the variation of the measurement unit 1 itself with respect to the characteristic value. Specifically, the measurement standard deviation S1 of the characteristic value is calculated from (Equation 4) using the standard deviation σEV of the measurement data of the characteristic value measured ten times for the same product 10 without changing the conditions. For example, S1 = 0.024242 pF is obtained. In the calculation of C4 (Equation 3), n is the number of repeated measurements ‘10’.

  The average value updating unit 206 determines whether the average value Xbar of the characteristic value calculated by the sample average calculating unit 202 is closer to the upper limit value SUL or the lower limit value SLL of the characteristic value for which the input is accepted, and corresponds to the closer one The upper limit value or lower limit value of the average value of the characteristic values of the confidence interval to be updated is updated as the average value Xbar of the characteristic values. For example, the average value Xbar of the characteristic values calculated by extracting 10 products as samples from the product lot in the sample average calculating means 202 and measuring by the measuring unit 1 is 2.5388 pF, so the upper limit value SUL of the characteristic values and The difference from the lower limit value SLL is calculated (Equation 5).

  Since (Equation 5) is close to the lower limit value SLL of the characteristic value, the lower limit value 2.525450 pF of the average value of the characteristic value of the 95% confidence interval calculated by the confidence interval calculation unit 205 is calculated as the average value of the characteristic values. Update as Xbar.

  The product standard deviation estimation means 207 estimates the standard deviation S2 of the characteristic value of the product itself based on the average value σsbar of the standard deviation of the characteristic value and the measured standard deviation S1 of the characteristic value. First, according to (Equation 6), the standard deviation TV of the characteristic value indicating the total variation is calculated.

  The control chart coefficients B4 and C4 in (Equation 6) are both calculated as B4 = 1.39558 and C4 = 0.99142 from the number of samples n = 30 for which the standard deviation for each product lot is obtained. By substituting this into (Equation 6), TV = 0.289008 pF is obtained.

  The standard deviation S2 of the characteristic value of the product itself can be calculated by (Equation 7). A standard deviation S2 = 0.0288137 pF of the product characteristic value is obtained.

  Since the average value Xbar of the product characteristic values, the standard deviation S2 of the product characteristic values, the upper limit value SUL and the lower limit value SLL of the product characteristic values are determined, the lower defect, which is the ratio of defective products below the lower limit value SLL The upper defective rate, which is the rate and the proportion of defective products exceeding the upper limit value SUL, can be calculated. The defect rate calculation means 208 is below the upper defect rate, which is the ratio of defective products exceeding the upper limit value of the predetermined characteristic value based on the product standard of the target product, and the lower limit value of the predetermined characteristic value based on the product standard of the target product. The lower defect rate, which is the ratio of defective products, is calculated.

  FIG. 4 is a schematic diagram showing an upper defect rate and a lower defect rate of the sampling data processing apparatus according to the first embodiment of the present invention. First, the SL value and Bias value shown in FIG. 4 are calculated. The SL value is a standardized distance from the standard center value of the characteristic value determined from the product standard to the upper limit value or the lower limit value, and the Bias value is a standardized distance from the average value of the product distribution characteristic value to the standard center value. Means the value.

  The SL value is calculated according to (Equation 8), and the Bias value is calculated according to (Equation 9).

  In (Expression 8), the standard deviation S2 = 0.0288137 of the characteristic value of the product itself, the upper limit value SUL = 2.64 of the characteristic value of the product, and the lower limit value SLL = 2.44, so that the SL value is 3.47058. . Further, in (Equation 9), the average value Xbar of the product characteristic value Xbar = 2.525450, the standard deviation S2 = 0.0288137 of the product itself, the upper limit value SUL = 2.64 of the product characteristic value, the lower limit value From SLL = 2.44, the Bias value is -0.50497.

  As shown in FIG. 4, the SL value and the Bias value are used to determine the probability of being equal to or lower than the upper limit value SUL of the predetermined characteristic value based on the product standard of the target product, and the lower limit of the predetermined characteristic value based on the product standard of the target product. A probability less than or equal to the value SLL can be calculated. That is, the value of the cumulative distribution function in the normal distribution having an average of 0 and a standard deviation of 1 is calculated using the standard density function f (z) in the normal distribution. The standard density function f (z) can be expressed by (Equation 10).

  First, calculation elements SS1 and SS2 are calculated using the SL value and the Bias value. Specifically, calculation elements SS1 and SS2 are calculated according to (Equation 11).

  Then, using the calculated calculation elements SS1 and SS2 as arguments, the probability PU below the upper limit value SUL of the predetermined characteristic value based on the product standard of the target product by the standard density function f (z) shown in (Expression 10), and the target A probability PL below a lower limit value SLL of a predetermined characteristic value based on the product standard of the product is calculated. Therefore, the upper defective rate 42, which is the ratio of defective products exceeding the upper limit value SUL of the predetermined characteristic value based on the product standard of the target product, can be obtained by (1-PU), and is determined based on the product standard of the target product. The lower defect rate 41, which is the ratio of defective products that are below the lower limit value SLL of the characteristic value, is the probability PL itself.

  In the above example, since the SL value is 3.47058 and the Bias value is −0.50497, SS1 = 3.9755, SS2 = −2.956561, and the upper defect rate 42 is 0.000035, the lower defect rate. 41 becomes 0.00151.

  Returning to FIG. 2, the non-defective product rate calculation unit 209 calculates the non-defective product rate based on the calculated upper defective rate 42 and lower defective rate 41. Specifically, the non-defective product rate can be calculated by (1− (upper defective rate + lower defective rate)), which is 0.998455. Note that the non-defective rate is “1” means that the non-defective rate is 100%.

  Therefore, it is possible to estimate the non-defective product rate with high accuracy from the sampling data of the product lot, and it is possible to estimate the quantity of product lots existing outside the product standard with high accuracy.

  FIG. 5 is a flowchart showing a processing procedure of the CPU 21 of the arithmetic processing unit 2 of the sampling data processing apparatus according to the first embodiment of the present invention. In FIG. 5, the CPU 21 of the arithmetic processing unit 2 accepts input of an upper limit value and a lower limit value of predetermined characteristic values based on the product standard of the target product (step S501).

  The CPU 21 acquires measurement data of a predetermined number of samples extracted from the product lot (step S502), and calculates an average value Xbar of characteristic values (step S503). The CPU 21 calculates an average value σsbar of standard deviations of characteristic values based on the product lot control chart (step S504).

  Based on the stored average value Xbar of the characteristic values and the calculated standard deviation average value σsbar, the CPU 21 determines the upper limit value and the lower limit value of the average value of the characteristic values in the 95% confidence interval. Calculate (step S505). CPU21 estimates measurement standard deviation S1 which shows the dispersion | variation in the measurement part 1 itself regarding a characteristic value (step S506).

  The CPU 21 calculates the difference between the average value Xbar of the characteristic values calculated and stored and the upper limit value SUL and the lower limit value SLL of the predetermined characteristic value for which the input has been received (step S507). It is determined whether or not the difference between the average value Xbar and the upper limit value SUL of the characteristic value is smaller (step S508). When the CPU 21 determines that the difference between the average characteristic value Xbar and the upper limit value SUL of the characteristic value is smaller (step S508: YES), the CPU 21 sets the upper limit value of the average characteristic value of the confidence interval as the characteristic. The average value Xbar is updated (step S509). When the CPU 21 determines that the difference between the average value Xbar of the characteristic values and the upper limit value SUL of the characteristic values is larger (step S508: NO), the CPU 21 sets the lower limit value of the average value of the characteristic values in the confidence interval as the characteristic value. The value is updated as the average value Xbar (step S510).

  The CPU 21 estimates the standard deviation S2 of the characteristic value of the product itself based on the average value σsbar of the standard deviation of the characteristic value and the measured standard deviation S1 of the characteristic value (step S511). The CPU 21 calculates an upper defective rate and a lower defective rate (step S512), and calculates a non-defective rate based on the calculated upper defective rate and lower defective rate (step S513).

  As described above, according to the first embodiment, the distribution of the product lot is specifically estimated by using the control limit of the control chart, and the worst value of the non-defective product rate is increased from the sampling data of the product lot. Since it can be estimated with accuracy, the quantity of product lots existing outside the product standard can be estimated with high accuracy.

(Embodiment 2)
Since the configuration of the sampling data processing apparatus according to the second embodiment of the present invention is the same as that of the first embodiment, detailed description thereof will be omitted by attaching the same reference numerals. The second embodiment is different from the first embodiment in that the minimum number of samples necessary for obtaining a predetermined non-defective product rate is calculated.

  FIG. 6 is a functional block diagram of the sampling data processing apparatus according to the second embodiment of the present invention. The measuring unit 1 measures a characteristic value indicating a predetermined characteristic of the product 10.

  The standard range receiving unit 201 receives an input of an upper limit value and a lower limit value of predetermined characteristic values based on the product standard of the target product. For example, assuming a product lot composed of 10,000 capacitors with a capacitance of 2.5 pF as a characteristic value, the upper limit product standard (upper limit value: SUL) of the capacitor is 2.75 pF, and the lower limit product standard (lower limit value). : SLL) is 2.26 pF. Therefore, the user inputs the upper limit value and the lower limit value of the predetermined characteristic value of the product via the keyboard 241 or the like. Of course, it may be stored in advance in the storage device 23 via a communication network (not shown).

  The standard deviation calculation means 203 calculates the average value of the standard deviation of the characteristic values based on the product lot control chart. As in the first embodiment, the average value 0.020531 pF of the standard deviation of the characteristic values of 30 lots is calculated as the average value σ sbar (pF) of the standard deviation of the entire characteristic values.

  The measurement standard deviation calculation means 204 estimates the measurement standard deviation S1 of the characteristic value indicating the variation of the measurement unit 1 itself with respect to the characteristic value. Specifically, the measurement standard deviation S1 of the characteristic value is calculated from (Equation 12) using the standard deviation σEV of the measurement data of the characteristic value obtained by measuring the same product 10 times without changing the conditions. For example, S1 = 0.024242 pF is obtained.

  The product standard deviation estimation means 207 estimates the standard deviation S2 of the characteristic value of the product itself based on the average value σsbar of the standard deviation of the characteristic value and the measured standard deviation S1 of the characteristic value. First, in accordance with (Equation 13), a standard deviation TV of characteristic values indicating total variation is calculated.

  The control chart coefficients B4 and C4 in (Equation 13) are both calculated as B4 = 1.39558 and C4 = 0.99142 from the number of samples n = 30 obtained for the standard deviation for each product lot. By substituting this into (Equation 13), TV = 0.289008 pF is obtained.

  The standard deviation S2 of the characteristic value of the product itself can be calculated by (Equation 14). A standard deviation S2 = 0.0288137 pF of the product characteristic value is obtained.

  The average specifying unit 210 sequentially specifies the average value of the characteristic values between the upper limit value and the lower limit value of the characteristic values received by the standard range receiving unit 201. Specifically, the variable i is sequentially incremented by “1” from 0 to 100, and the average value Xbar of the product characteristic value is calculated according to (Equation 15), and specified as the average value of the characteristic value.

  For example, in (Expression 15), when the variable i = 0, Xbar = SLL = 2.26 pF. The defect rate calculation means 211 is an upper defect that is a ratio of defective products that sequentially exceed the upper limit value of the characteristic value based on the product standard of the target product for each average value Xbar of the characteristic values specified while sequentially changing the variable i. The lower defect rate, which is the ratio of defective products that fall below the lower limit value of the characteristic value based on the rate and the product standard of the target product, is calculated.

  As in the first embodiment, the SL value is calculated according to (Expression 16), and the Bias value is calculated according to (Expression 17).

  Then, calculation elements SS1 and SS2 are first calculated using the SL value and the Bias value. Specifically, calculation elements SS1 and SS2 are calculated according to (Equation 18).

  Then, using the calculated calculation elements SS1 and SS2 as arguments, the probability PU below the upper limit value SUL of the predetermined characteristic value based on the product standard of the target product by the standard density function f (z) shown in (Expression 10), and the target A probability PL below a lower limit value SLL of a predetermined characteristic value based on the product standard of the product is calculated. Therefore, as in the first embodiment, in FIG. 4, the upper defect rate 42, which is the proportion of defective products exceeding the upper limit value SUL of the predetermined characteristic value based on the product standard of the target product, is obtained by (1−PU). The lower defect rate 41, which is the proportion of defective products that are below the lower limit value SLL of the predetermined characteristic value based on the product standard of the target product, is the probability PL itself. When the variable i = 0, the upper defect rate 42 is 0.0 and the lower defect rate 41 is 0.5.

  Returning to FIG. 6, the non-defective product rate calculation unit 212 sequentially calculates the non-defective product rate based on the calculated upper defect rate 42 and lower defect rate 41. Specifically, the non-defective product rate can be calculated by (1− (upper defective rate + lower defective rate)), which is 0.5. All non-defective product rates are calculated while varying the variable i from 0 to 100.

  The sample number specifying means 213 sequentially specifies the average value section of the characteristic values in which the calculated non-defective product ratio is equal to or greater than a predetermined threshold, and calculates the sampling number of the corresponding sample. FIG. 7 is a graph showing the variation of the yield rate when the variable i is varied from 0 to 100. In FIG. 7, the vertical axis represents the yield rate, and the horizontal axis represents the average value Xbar of the characteristic values from the lower limit value SLL to the upper limit value SUL. In FIG. 7, for example, a non-defective product rate of “1” means that the non-defective product rate is 100%.

  FIG. 8 is a graph in which the two areas 71 and 72 in FIG. 7 are enlarged and displayed. For example, when obtaining the section of the average value Xbar of the characteristic values at which the non-defective product ratio is 99.99% or more, the section having the non-defective product ratio of 99.99% is from 2.3678 pF to FIG. 8A and FIG. Since it is 2.6422 pF, the length of the section is 0.2744 pF. An allowable range of section variation is set above and below this section. As is clear from FIGS. 7, 8 (a) and 8 (b), the allowable width of the section variation is set at two places in the section where the non-defective product ratio is 99.99% or more. In the first embodiment, the allowable width of the section variation is set to 0.015 pF. Further, using the average value σ sbar = 0.020531 pF of the standard deviation of the characteristic values of the entire product calculated by the standard deviation calculating means 203, (Equation 19) is established.

  By calculating back (Equation 19), the sample sampling number n = 13.3 can be obtained, and the fractional part can be rounded up to obtain n = 14.

  FIG. 9 is a flowchart showing a processing procedure of the CPU 21 of the arithmetic processing unit 2 of the sampling data processing apparatus according to the second embodiment of the present invention. In FIG. 9, the CPU 21 of the arithmetic processing unit 2 accepts input of an upper limit value and a lower limit value of a predetermined characteristic value based on the product standard of the target product (step S901).

  The CPU 21 calculates the average value σsbar of the standard deviation of the characteristic values based on the product lot control chart (step S902). CPU21 estimates measurement standard deviation S1 which shows the dispersion | variation in the measurement part 1 itself regarding a characteristic value (step S903).

  The CPU 21 estimates the standard deviation S2 of the characteristic value of the product itself based on the average value σsbar of the standard deviation of the characteristic value and the measured standard deviation S1 of the characteristic value (step S904). The CPU 21 sets the variable i to “0” (step S905), and sequentially specifies the average value Xbar of the characteristic values between the upper limit value SUL and the lower limit value SLL of the predetermined characteristic value that has received the input (step S906). ).

  The CPU 21 sequentially calculates the upper defect rate and the lower defect rate for each specified average value of the characteristic values (step S907), and sequentially determines the non-defective rate based on the calculated upper defect rate and lower defect rate. Is calculated (step S908). Specifically, the non-defective product rate can be calculated by (1− (upper defective rate + lower defective rate)).

  The CPU 21 determines whether or not the variable i exceeds “100” (step S909). When the CPU 21 determines that the variable i does not exceed “100” (step S909: NO), the CPU 21 increments the variable i by “1” (step S910), returns the process to step S906, and described above. Repeat the process.

  When the CPU 21 determines that the variable i has exceeded “100” (step S909: YES), the CPU 21 calculates the sampling number n (step S911).

  As described above, according to the second embodiment, by using the control chart data, the distribution of the product lot is specifically estimated, and by estimating the distribution of the non-defective product ratio, a predetermined non-defective product rate, For example, it is possible to reversely calculate the number of samples to be extracted to be 99.99% or more.

  The present invention is not limited to the above-described embodiments, and various changes and improvements can be made within the scope of the present invention. For example, in Embodiments 1 and 2 described above, the product distribution is estimated as a normal distribution, but it goes without saying that the present invention can be applied even if the product distribution is an arbitrary distribution other than the normal distribution. .

1 Measurement Unit 2 Arithmetic Processing Unit 10 Product 21 CPU
22 memory 23 storage device 24 I / O interface 25 video interface 26 portable disk drive 27 measurement interface 28 internal bus 90 portable recording medium 230 computer program 241 keyboard 242 mouse 251 display device

Claims (6)

In the sampling data processing device that estimates the quantity of product lots that are outside the product standard from the sampling data of the product lot of the target product,
A standard range receiving means for receiving an input of an upper limit value and a lower limit value of a predetermined characteristic value based on the product standard of the target product;
Standard deviation calculating means for calculating an average value of standard deviations of the characteristic values based on the control chart of the product lot;
Sample average calculation means for calculating an average value of the characteristic values from measurement data of a predetermined number of samples extracted from the product lot;
A confidence interval calculation means for calculating an upper limit value and a lower limit value of the average value of the characteristic value of the confidence interval of 95% reliability based on the calculated average value of the standard deviation of the characteristic value and the average value of the characteristic value; ,
A measurement standard deviation calculating means for estimating a measurement standard deviation indicating variation of the measuring device itself with respect to the characteristic value;
It is determined whether the calculated average value of the characteristic values is closer to the upper limit value or the lower limit value of the characteristic values for which input has been accepted, and the upper limit value of the average value of the characteristic values of the confidence interval corresponding to the closer one Or an average value updating means for updating a lower limit value as an average value of the characteristic values;
Product standard deviation estimating means for estimating the standard deviation of the characteristic value of the product itself based on the average value of the standard deviation of the characteristic value and the measurement standard deviation;
The upper defective rate, which is the proportion of defective products that exceed the upper limit value of the predetermined characteristic value based on the product standard of the target product, and the ratio of defective products that are lower than the lower limit value of the predetermined characteristic value based on the product standard of the target product A defect rate calculating means for calculating a side defect rate;
A sampling data processing apparatus comprising: a non-defective product rate calculating unit that calculates a non-defective product rate based on the calculated upper defective rate and the lower defective rate. In the sampling data processing device that estimates the quantity of product lots that are outside the product standard from the sampling data of the product lot of the target product,
A standard range receiving means for receiving an input of an upper limit value and a lower limit value of a predetermined characteristic value based on the product standard of the target product;
Standard deviation calculating means for calculating an average value of standard deviations of the characteristic values based on the control chart of the product lot;
A measurement standard deviation calculating means for estimating a measurement standard deviation indicating variation of the measuring device itself with respect to the characteristic value;
Product standard deviation estimating means for estimating the standard deviation of the characteristic value of the product itself based on the average value of the standard deviation of the characteristic value and the measurement standard deviation;
Average specifying means for sequentially specifying an average value of the characteristic values between the upper limit value and the lower limit value of the characteristic values received from the input;
For each specified average value of the characteristic values, an upper defective rate that is a ratio of defective products exceeding an upper limit value of a predetermined characteristic value based on the product standard of the target product, and a predetermined value based on the product standard of the target product A defective rate calculating means for calculating a lower defective rate which is a ratio of defective products below a lower limit value of the characteristic value;
A non-defective product rate calculating means for sequentially calculating a non-defective product rate based on the calculated upper defective rate and the lower defective rate,
A sampling data processing apparatus characterized by sequentially calculating the number of samplings of the corresponding sample by specifying a section of the average value of the characteristic values in which the calculated non-defective product rate is equal to or greater than a predetermined threshold value. In a sampling data processing method that can be executed by a sampling data processing device that estimates the quantity of a product lot that exists outside the product standard from the sampling data of the product lot of the target product,
The sampling data processing device includes:
Receiving an input of an upper limit value and a lower limit value of a predetermined characteristic value based on the product standard of the target product;
Calculating an average value of the standard deviation of the characteristic values based on the control chart of the product lot;
Calculating the average value of the characteristic values from the measurement data of a predetermined number of samples extracted from the product lot;
Calculating an upper limit value and a lower limit value of the average value of the characteristic value of the 95% confidence interval based on the calculated average value of the standard deviation of the characteristic value and the average value of the characteristic value;
Estimating a measurement standard deviation indicating variation of the measuring device itself with respect to the characteristic value;
It is determined whether the calculated average value of the characteristic values is closer to the upper limit value or the lower limit value of the characteristic values for which input has been accepted, and the upper limit value of the average value of the characteristic values of the confidence interval corresponding to the closer one Or a step of updating a lower limit value as an average value of the characteristic values;
Estimating the standard deviation of the characteristic value of the product itself based on the average value of the standard deviation of the characteristic value and the measurement standard deviation;
The upper defective rate, which is the proportion of defective products that exceed the upper limit value of the predetermined characteristic value based on the product standard of the target product, and the ratio of defective products that are lower than the lower limit value of the predetermined characteristic value based on the product standard of the target product Calculating a side defect rate;
And a step of calculating a non-defective product rate based on the calculated upper defective rate and lower lower defective rate. In a sampling data processing method that can be executed by a sampling data processing device that estimates the quantity of a product lot that exists outside the product standard from the sampling data of the product lot of the target product,
The sampling data processing device includes:
Receiving an input of an upper limit value and a lower limit value of a predetermined characteristic value based on the product standard of the target product;
Calculating an average value of the standard deviation of the characteristic values based on the control chart of the product lot;
Estimating a measurement standard deviation indicating variation of the measuring device itself with respect to the characteristic value;
Estimating the standard deviation of the characteristic value of the product itself based on the average value of the standard deviation of the characteristic value and the measurement standard deviation;
Between the upper limit value and the lower limit value of the characteristic value that has received an input, sequentially specifying an average value of the characteristic value;
For each specified average value of the characteristic values, an upper defective rate that is a ratio of defective products exceeding an upper limit value of a predetermined characteristic value based on the product standard of the target product, and a predetermined value based on the product standard of the target product A step of calculating a lower defect rate that is a ratio of defective products below a lower limit value of the characteristic value;
And sequentially calculating a non-defective product rate based on the calculated upper defect rate and lower defect rate,
A sampling data processing method characterized by sequentially calculating the sampling number of the corresponding sample by specifying a section of the average value of the characteristic values in which the calculated non-defective product rate is equal to or greater than a predetermined threshold value. In a computer program that can be executed by a sampling data processing device that estimates the quantity of product lots that are outside the product standard from sampling data of the product lot of the target product,
The sampling data processing device;
Standard range receiving means for receiving an input of an upper limit value and a lower limit value of a predetermined characteristic value based on the product standard of the target product,
A standard deviation calculating means for calculating an average value of standard deviations of the characteristic values based on the control chart of the product lot;
Sample average calculating means for calculating an average value of the characteristic values from measurement data of a predetermined number of samples extracted from the product lot,
A confidence interval calculation means for calculating an upper limit value and a lower limit value of the average value of the characteristic value of the confidence interval of 95% reliability, based on the calculated average value of the standard deviation of the characteristic value and the average value of the characteristic value;
A measurement standard deviation calculating means for estimating a measurement standard deviation indicating variation of the measuring device itself with respect to the characteristic value;
It is determined whether the calculated average value of the characteristic values is closer to the upper limit value or the lower limit value of the characteristic values for which input has been accepted, and the upper limit value of the average value of the characteristic values of the confidence interval corresponding to the closer one Or an average value updating means for updating a lower limit value as an average value of the characteristic values,
Product standard deviation estimating means for estimating the standard deviation of the characteristic value of the product itself based on the average value of the standard deviation of the characteristic value and the measured standard deviation;
The upper defective rate, which is the proportion of defective products that exceed the upper limit value of the predetermined characteristic value based on the product standard of the target product, and the ratio of defective products that are lower than the lower limit value of the predetermined characteristic value based on the product standard of the target product A computer program that functions as a defect rate calculation unit that calculates a side defect rate, and a non-defective rate calculation unit that calculates a non-defective rate based on the calculated upper defect rate and lower defect rate. In a computer program that can be executed by a sampling data processing device that estimates the quantity of product lots that are outside the product standard from sampling data of the product lot of the target product,
The sampling data processing device;
Standard range receiving means for receiving an input of an upper limit value and a lower limit value of a predetermined characteristic value based on the product standard of the target product,
A standard deviation calculating means for calculating an average value of standard deviations of the characteristic values based on the control chart of the product lot;
A measurement standard deviation calculating means for estimating a measurement standard deviation indicating variation of the measuring device itself with respect to the characteristic value;
Product standard deviation estimating means for estimating the standard deviation of the characteristic value of the product itself based on the average value of the standard deviation of the characteristic value and the measured standard deviation;
Average specifying means for sequentially specifying an average value of the characteristic values between an upper limit value and a lower limit value of the characteristic values that have received the input,
For each specified average value of the characteristic values, an upper defective rate that is a ratio of defective products exceeding an upper limit value of a predetermined characteristic value based on the product standard of the target product, and a predetermined value based on the product standard of the target product A defective rate calculation means for calculating a lower defective rate that is a ratio of defective products that are lower than a lower limit value of the characteristic value;
Based on the calculated upper defective rate and lower lower defective rate, a non-defective product rate calculating means for sequentially calculating a non-defective product rate, and an average value of the characteristic values in which the calculated non-defective product rate is equal to or greater than a predetermined threshold. A computer program that functions as means for calculating the number of samples taken by specifying a section.